In this paper, we present our experiment and simulation study of a free piston linear engine based on operating conditions and structure of the linear engine for generating electric power. The free piston linear engine includes a two-stroke free piston engine, linear generators, and compressors. In the experimental study, the effects of key parameters such as input caloric value, equivalence ratio, spark timing delay, electrical resistance, and air gap length on the piston dynamics and electric power output are investigated. Propane is used as a fuel in the free piston linear engine, and it is premixed with the air to make a homogeneous charge before go into the cylinder. The air and fuel mass flow rate are varied by a mass flow controller. The experimental results show that the maximum generating power is found with the value of 111 W at the input caloric value of 5.88 kJ/s, spark timing delay of 1.5 ms, equivalence ratio of 1.0, electric resistance of 30 Omega, and air gap length of 1.0 mm. In order to check the durability of the linear engine, a durable test is conducted during 100 h. The experimental results show that there are no problems for the linear engine after about one hundred hours of the durable test. Beside experimental study, a simulation study is conducted to predict operating behavior of the linear engine. In the simulation study, the two-stroke free piston linear engine is modeled and simulated through a combination of three mathematical models including a dynamic model, a linear alternator model and a thermodynamic model. These mathematical models are combined and solved by a program written in Fortran. Besides, the effects of key parameters such as reciprocating mass, spark timing and spring stiffness on the piston dynamics and electric power output of the linear engine are also investigated. The simulation results show that the simulation and experimental data are nearly similar at the same initial conditions. In addition, a highest generating power of the linear engine can be easily found by optimizing the key parameters. (C) 2015 Elsevier Ltd. All rights reserved.